Finance and Financial Management Commons^™

Shrinking Factor Dimension: A Reduced-Rank Approach, Ai He, Dashan Huang, Jiaen Li, Guofu Zhou

Research Collection Lee Kong Chian School Of Business

We propose a reduced-rank approach (RRA) to reduce a large number of factors to a few parsimonious ones. In contrast to PCA and PLS, the RRA factors are designed to explain the cross section of stock returns, not to maximize factor variations or factor covariances with returns. Out of 70 factor proxies, we find that five RRA factors outperform the Fama-French (2015) five factors for pricing target portfolios, but performs similarly for pricing individual stocks. Our results suggest that existing factor proxies do not provide enough new information at the stock level beyond the Fama-French (2015) five factors.